CODEWITHSUNDEEP
pythonpython-3.xnumpyrandom
PRO88
PRO88

Reputation: 437

Generate numpy array of random integers within a range and some fixed integers outside the defined range

I want to combine two random number generator.

I want to generate size = 20 random integers between [-5,5] and a fixed integer -999.

I tried following code:

from random import randint, choice  
import numpy as np
np.random.seed(1)
dd = np.array([choice([randint(-999,-999),randint(-5,5)]) for _ in range(20)])
print(dd)

Result:

[-999 -999 -999   -4 -999 -999   -4 -999 -999    3    5   -3    2    0
   -1 -999 -999   -5 -999   -4]

Is there any better way to generate random integers? The result with current code has many -999 due to same upper and lower limit.

Upvotes: 0

Views: 968

Answers (2)

agus.dinamarca
agus.dinamarca

Reputation: 99

Your solution yields approximately 50% of the cases -999 and the rest a number in [-5, 5]. This is because choice has to choose between -999 and X (where X is a number previously defined by randint, between -5 and 5). If you want every number to have the same probability of being chosen try the following proposed solution:

lst = [i for i in range(-5, 6)] + [-999]
dd = [choice(lst) for _ in range(20)]

The probability that a number in the list lst is chosen by choice is 1/12 ~ 8%

Upvotes: 1

Tee Yee Yang
Tee Yee Yang

Reputation: 11

Currently its choosing between -999 and random number between -5 and 5 on each iteration. So you would expect 50% of results to be -999.

One way would be to first define a list of choices

choices = list(range(-5,5))+ [-999]

Then, do the sampling using np.random.choice instead of a for loop list comprehension. Using size = 20 you can set how many to select.

dd = np.random.choice(choices, size = 20)

Upvotes: 0

Related Questions